INDEX

Explanations

opening quotation marks

This neuron activates on numeric values (especially decimals) in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

是非

-1.29

のご紹介

-1.27

哈哈哈哈

-1.26

 GONNA

-1.23

になっています

-1.23

 PLAYING

-1.22

vay

-1.21

超市

-1.20

یات

-1.19

glades

-1.19

POSITIVE LOGITS

“

2.25

「

1.44

 those

1.44

etlen

1.43

norme

1.42

 auténtico

1.42

 starken

1.39

 resistentes

1.38

 zondag

1.38

 cuyos

1.37

Activations Density 0.006%